A liveness testing apparatus includes a testing circuit. The testing circuit is configured to test a liveness of an object included in a received input image based on whether an image of the object has a characteristic indicative of a flat surface or a characteristic indicative of a three-dimensional (3D) structure.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A liveness testing method for an electronic device, the method comprising: testing a liveness of an object included in a received input image based on whether an image of the object has a characteristic indicative of a flat surface or a characteristic indicative of a three-dimensional (3D) structure, wherein the testing includes calculating diffusion speeds for a plurality of pixels corresponding to the image of the object based on values of the plurality of pixels before diffusion and diffusion values of the plurality of pixels after diffusion, and determining whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on the diffusion speeds for the plurality of pixels corresponding to the image of the object.
A liveness detection method for a device analyzes an input image to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process. These diffusion speeds are then used to distinguish between flat surfaces and 3D structures, indicating whether the object is live or not.
2. The method of claim 1 , wherein the image of the object included in the received input image corresponds to a face.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The object in the input image is specifically a face.
3. The method of claim 1 , wherein the determining comprises: determining whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on a distribution of light energy included in the plurality of pixels corresponding to the image of the object.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The method determines if the object is flat or 3D based on the distribution of light energy within the pixels of the object's image.
4. The method of claim 1 , wherein the determining comprises: determining whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on a degree of uniformity in a distribution of light energy included in the plurality of pixels corresponding to the image of the object.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The method determines if the object is flat or 3D based on the uniformity of light energy distribution across the object's pixels.
5. The method of claim 1 , wherein the determining comprises: determining whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on statistical information related to the diffusion speeds.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The determination of flatness or 3D structure relies on statistical information derived from these diffusion speeds.
6. The method of claim 5 , wherein the testing further comprises: iteratively calculating values of the plurality of pixels based on a diffusion equation; and wherein the calculating calculates the diffusion speeds for each of the plurality of pixels based on a difference between a pixel value before each iterative calculation and a pixel value after each iterative calculation.
The liveness detection method for a device that analyzes an input image to determine if an object is live and determines if the object is flat or 3D based on statistical information derived from diffusion speeds as described, includes iteratively refining pixel values using a diffusion equation. The "diffusion speeds" are calculated from the difference in pixel values before and after each iteration of the diffusion equation.
7. The method of claim 5 , wherein the statistical information related to the diffusion speeds comprises at least one of: a number of pixels, among the plurality of pixels, having diffusion speeds greater than or equal to a threshold value; a distribution of the pixels having the diffusion speeds greater than or equal to the threshold value; an amount of noise components included in a first scale region extracted based on magnitudes of the diffusion speeds; an average of the diffusion speeds; a standard deviation of the diffusion speeds; and a filter response based on the diffusion speeds.
The liveness detection method for a device that analyzes an input image to determine if an object is live and determines if the object is flat or 3D based on statistical information derived from diffusion speeds as described where the determination of flatness or 3D structure relies on statistical information derived from these diffusion speeds. The statistical information includes: number of pixels with diffusion speeds above a threshold; the spatial distribution of these pixels; the amount of noise in a region extracted based on diffusion speed magnitudes; the average diffusion speed; the standard deviation of diffusion speeds; and a filter response based on the diffusion speeds.
8. The method of claim 1 , wherein the testing further comprises: filtering the received input image to generate a filtered image; and wherein the determining determines whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on statistical information related to a change in values of the plurality of pixels corresponding to the image of the object in the received input image and a plurality of pixels corresponding to the image of the object in the filtered image.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The method first filters the input image and then compares statistical properties of pixel values between the original and filtered images to determine if the object is flat or 3D.
9. The method of claim 8 , wherein the filtering comprises: diffusing the received input image to generate a diffusion image; and wherein the calculating calculates the diffusion speeds for each of the plurality of pixels corresponding to the image of the object based on a difference between a value of each pixel in the received input image and a corresponding value of the pixel in the diffusion image.
The liveness detection method for a device that analyzes an input image to determine if an object is live and determines if the object is flat or 3D based on statistical properties of pixel values between the original and filtered images as described, uses a diffusion process to generate the filtered image. The "diffusion speeds" are calculated as the difference between the pixel values in the original image and the corresponding pixel values in the diffused image.
10. The method of claim 1 , further comprising at least one of: outputting a signal corresponding to a failed test when the object is determined to have the characteristic indicative of the flat surface; and outputting a signal corresponding to a successful test when the object is determined to have the characteristic indicative of the 3D structure.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The method outputs a "failed" signal if a flat surface is detected, or a "success" signal if a 3D structure is detected.
11. The method of claim 1 , wherein the input image is a single image.
The liveness detection method for a device as described where an input image is analyzed to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process, and uses these diffusion speeds to distinguish between flat surfaces and 3D structures. The input image is a single still image.
12. A liveness testing method comprising: filtering a received image including an image of an object to generate a filtered image, the filtering including diffusing pixels corresponding to the image of the object in the received image to generate the filtered image; determining a magnitude of change in values of the pixels corresponding to the image of the object in the filtered image and the received image; calculating diffusion speeds for the pixels corresponding to the image of the object based on the values of the pixels corresponding to the image of the object in the received image and the filtered image; and testing a liveness of the object based on the diffusions speeds.
A liveness detection method involves filtering a received image containing an object to generate a filtered image by diffusing the pixels. The method then determines the magnitude of change in pixel values between the filtered and original images, calculating "diffusion speeds" based on pixel values in both images. The liveness of the object is tested using these diffusion speeds.
13. The method of claim 12 , wherein the object corresponds to a face.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, where the object being tested is a face.
14. The method of claim 12 , wherein the diffusing comprises: iteratively updating values of the pixels based on a diffusion equation.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, where diffusing the pixels involves iteratively updating pixel values according to a diffusion equation.
15. The method of claim 14 , wherein the iteratively updating iteratively updates the values of the pixels by applying an additive operator splitting (AOS) scheme to the diffusion equation.
The liveness detection method using image filtering, pixel diffusion with iterative updating via a diffusion equation, and diffusion speeds to test object liveness, as described, where an Additive Operator Splitting (AOS) scheme is applied to the diffusion equation during the iterative updating process.
16. The method of claim 12 , wherein the testing comprises: estimating a surface property related to the object based on the diffusion speeds; and testing the liveness of the object based on the estimated surface property.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, includes estimating a surface property of the object based on the calculated diffusion speeds, and then using this estimated surface property to determine liveness.
17. The method of claim 16 , wherein the surface property comprises at least one of: a light-reflective property of a surface of the object; a number of dimensions of the surface of the object; and a material of the surface of the object.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, includes estimating a surface property of the object based on the calculated diffusion speeds, and then using this estimated surface property to determine liveness, as described, where the surface property includes the light reflectivity, dimensionality (2D or 3D), and/or material of the object's surface.
18. The method of claim 16 , wherein the estimating comprises: analyzing a distribution of light energy included in the image of the object based on the diffusion speeds to estimate the surface property.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, includes estimating a surface property of the object based on the calculated diffusion speeds, and then using this estimated surface property to determine liveness, as described, where the estimation involves analyzing the distribution of light energy in the image based on the diffusion speeds.
19. The method of claim 16 , further comprising at least one of: outputting a signal corresponding to a failed test when the estimated surface property corresponds to a surface property of a medium displaying a face; and outputting a signal corresponding to a successful test when the estimated surface property corresponds to a surface property of an actual face.
The liveness detection method using image filtering, pixel diffusion, diffusion speeds, and surface property estimation to test object liveness, includes estimating a surface property of the object based on the calculated diffusion speeds, and then using this estimated surface property to determine liveness, as described, where a "failed" signal is output if the estimated surface property matches that of a medium displaying a face (e.g., a screen), and a "success" signal is output if it matches that of an actual face.
20. The method of claim 12 , wherein the testing comprises: calculating statistical information related to the diffusion speeds; and testing the liveness of the object based on the calculated statistical information.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, involves calculating statistical information from the diffusion speeds and using this statistical information to determine liveness.
21. The method of claim 20 , wherein the calculating statistical information comprises at least one of: calculating a number of pixels having diffusion speeds greater than or equal to a threshold value, among the diffusion speeds; calculating a distribution of the pixels having the diffusion speeds greater than or equal to the threshold value, among the diffusion speeds; calculating at least one of an average and a standard deviation of the diffusion speeds; and calculating a filter response based on the diffusion speeds.
The liveness detection method using image filtering, pixel diffusion, diffusion speeds, and statistical information to test object liveness, as described, where the statistical information includes: the number of pixels with diffusion speeds exceeding a threshold; the distribution of those pixels; the average and/or standard deviation of the diffusion speeds; and a filter response calculated using the diffusion speeds.
22. The method of claim 20 , wherein the calculating statistical information comprises: extracting a first scale region from the received image based on magnitudes of the diffusion speeds; and extracting a characteristic of the first scale region; and wherein the testing tests the liveness of the object based on the extracted characteristic.
The liveness detection method using image filtering, pixel diffusion, diffusion speeds, and statistical information to test object liveness, as described, involves extracting a region of interest based on diffusion speed magnitudes, analyzing a characteristic of this region, and using that characteristic to test liveness.
23. The method of claim 22 , wherein the characteristic of the first scale region includes an amount of noise components included in the first scale region; and the noise components are calculated based on a difference between the first scale region and a result of applying median filtering to the first scale region.
The liveness detection method that uses image filtering, pixel diffusion, diffusion speeds, statistical information extracted from a region of interest, and tests object liveness as described, where the characteristic of the region of interest is the amount of noise, calculated by comparing the region to a median-filtered version of itself.
24. The method of claim 20 , further comprising: outputting a signal corresponding to a failed test when the statistical information corresponds to statistical information related to a medium displaying a face; and outputting a signal corresponding to a successful test when the statistical information corresponds to statistical information related to an actual face.
The liveness detection method using image filtering, pixel diffusion, diffusion speeds, statistical information, and tests object liveness as described outputs a "failed" signal if the statistical information matches that of a displayed face, and a "success" signal if it matches that of an actual face.
25. The method of claim 12 , wherein the calculating the diffusion speeds comprises: calculating a diffusion speed for each of the pixels based on an original value of the pixel before diffusion and a diffusion value of the pixel after diffusion.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, calculates the diffusion speed for each pixel based on its original value before diffusion and its value after diffusion.
26. The method of claim 25 , wherein the calculated diffusion speed of the pixel increases as the difference between the original value and the diffusion value increases, and the calculated diffusion speed of the pixel decreases as the difference between the original value and the diffusion value decreases.
The liveness detection method calculates the diffusion speed for each pixel based on its original value before diffusion and its value after diffusion, and uses image filtering, pixel diffusion, and diffusion speeds to test object liveness as described. The calculated diffusion speed increases as the difference between the original and diffused values increases, and decreases as the difference decreases.
27. The method of claim 12 , wherein the input image corresponds to a single image of a face of a user.
The liveness detection method using image filtering, pixel diffusion, and diffusion speeds to test object liveness, as described, where the input image is a single image of a user's face.
28. A non-transitory computer-readable medium comprising a program that, when executed on a computer device, causes the computer device to perform the method of claim 1 .
A non-transitory computer-readable storage medium storing instructions that, when executed, perform the steps of the liveness detection method for a device analyzes an input image to determine if an object is a live, three-dimensional object or a flat, printed image. The method calculates "diffusion speeds" for pixels in the image based on pixel values before and after a diffusion process. These diffusion speeds are then used to distinguish between flat surfaces and 3D structures, indicating whether the object is live or not.
29. A liveness testing apparatus comprising: a testing circuit configured to test a liveness of an object included in a received input image based on whether an image of the object has a characteristic indicative of a flat surface or a characteristic indicative of a three-dimensional (3D) structure, wherein the testing circuit is further configured to calculate diffusion speeds for a plurality of pixels corresponding to the image of the object based on values of the plurality of pixels before diffusion and diffusion values of the plurality of pixels after diffusion, and determine whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on the diffusion speeds for the plurality of pixels corresponding to the image of the object.
A liveness detection apparatus includes a circuit that analyzes an input image to determine if an object is live based on whether the image exhibits characteristics of a flat surface or a 3D structure. The circuit calculates "diffusion speeds" for the image's pixels based on pixel values before and after diffusion, using these speeds to differentiate between flat and 3D objects.
30. The apparatus of claim 29 , wherein the object corresponds to a face.
The liveness detection apparatus includes a circuit that analyzes an input image to determine if an object is live based on the surface characteristic (flat or 3D). The circuit calculates "diffusion speeds" for the image's pixels. The object being analyzed is a face.
31. The apparatus of claim 29 , wherein the testing circuit is further configured to, filter the received input image to generate a filtered image, and determine whether the image of the object has the characteristic indicative of the flat surface or the characteristic indicative of the 3D structure based on statistical information related to a change in values of the plurality of pixels corresponding to the image of the object in the received input image and a plurality of pixels corresponding to the image of the object in the filtered image.
The liveness detection apparatus includes a circuit that analyzes an input image to determine if an object is live based on the surface characteristic (flat or 3D). The circuit calculates "diffusion speeds" for the image's pixels. The circuit first filters the input image and then determines liveness based on statistical differences between pixel values in the original and filtered images.
32. The apparatus of claim 31 , wherein the testing circuit is further configured to, filter the received input image by diffusing the received input image to generate a diffusion image, calculate the diffusion speeds for each of the plurality of pixels corresponding to the image of the object based on a difference between a value of each pixel in the received input image and a corresponding value of the pixel in the diffusion image.
The liveness detection apparatus includes a circuit that analyzes an input image to determine if an object is live based on statistical differences between pixel values in the original and filtered images. The circuit first filters the input image using diffusion. The "diffusion speeds" are then calculated as the difference between pixel values in the original image and the corresponding values in the diffused image.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
February 3, 2015
June 13, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.